MEMRISTOR-BASED NAIVE BAYES CLASSIFIER DESIGN METHOD AND SYSTEM, AND MEMRISTOR-BASED NAIVE BAYES CLASSIFIER

A memristor-based naive Bayes classifier design method and system, and a memristor-based naive Bayes classifier, belonging to the technical field of information. The method comprises: constructing a naive Bayes classifier, which includes a memristor array having M rows x 2N columns, wherein M is the number of classification types, and N is the number of pixel points in a picture (S1); calculating the number hj,2i-1 of pixel values, which are 0, in an ith pixel point in a jth training sample and the number hj,2i of pixel values, which are 1, in the ith pixel point in the jth training sample, wherein j = 1, 2, ..., M (S2); and applying hj,2i-1 pulses to a memristor Rj,2i-1 in a jth row and a (2i-1)th column, so as to modulate the conductance of Rj,2i-1, and applying hj,2i pulses to a memristor Rj,2i in the jth row and a 2ith column, so as to modulate the conductance of Rj,2i, wherein a modulated naive Bayes classifier is used for identifying the type of a picture to be classified (S3). Asymmetrical and nonlinea